From 6fe7c948e1bc937b99f73866e2bc2cb27d7cc3fd Mon Sep 17 00:00:00 2001 From: Gustaf Rydholm Date: Thu, 9 May 2024 13:36:23 +0200 Subject: Restructure --- content/cv.md | 30 +++++++++++++++--------------- 1 file changed, 15 insertions(+), 15 deletions(-) (limited to 'content') diff --git a/content/cv.md b/content/cv.md index f26312d..3b6bc38 100644 --- a/content/cv.md +++ b/content/cv.md @@ -82,6 +82,21 @@ be a better alternative of k8s. I am also moving to [podman](https://podman.io/) next time I work with containers in my spare time. I am also keen on working more with service meshes, to improve my knowledge of their capabilities. +#### Machine Learning + +I have worked on many machine learning projects, both at work and at home (hobby), ranging +from time series classification, regression, generative models, and machine translation to +name a few. My latest [project](https://github.com/aktersnurra/rag) is a Retrieval +Augmented Generation (RAG) system for my personal library of books and articles. I plan +on extending this to integrate with dmenu, so I can ask questions through the search bar, +which will be pretty cool. + +I have also spent quite some time on a OCR [project](https://github.com/aktersnurra/text-recognizer) +where I started with CNN-LSTM based systems to then move on to transformer models that +I have implemented. This has been a crucial project for me, as it has kept me up to date +with the SOTA in transformer architectures, and allowed me to test some of my own ideas out, +most of which were not successful. + #### Operating Systems I use Artix Linux as my daily OS. However, I would like to transition to OpenBSD @@ -131,21 +146,6 @@ extracted into separate services if necessary. But I see a downside to this appr and that is how to deal with the complexity of testing multiple features/branches at the same time in dev/staging, maybe there is a good way? -#### Machine Learning - -I have worked on many machine learning projects, both at work and at home (hobby), ranging -from time series classification, regression, generative models, and machine translation to -name a few. My latest [project](https://github.com/aktersnurra/rag) is a Retrieval -Augmented Generation (RAG) system for my personal library of books and articles. I plan -on extending this to integrate with dmenu, so I can ask questions through the search bar, -which will be pretty cool. - -I have also spent quite some time on a OCR [project](https://github.com/aktersnurra/text-recognizer) -where I started with CNN-LSTM based systems to then move on to transformer models that -I have implemented. This has been a crucial project for me, as it has kept me up to date -with the SOTA in transformer architectures, and allowed me to test some of my own ideas out, -most of which were not successful. - #### Miscellaneous | What | Level | -- cgit v1.2.3-70-g09d2